4 ms·

If your home pages looks like this (https://www.ycombinator.com https://www.ycombinator.com), you can expect people to try want to "launch" their startups on yo

If your home pages looks like this (https://www.ycombinator.com https://www.ycombinator.com), you can expect people to try want to "launch" their startups on your forum. There seems to be a large contingent of HN users that have settled on the belief that if something is done using AI it is "lazy slop" and it needs to be shunned! Obviously this can be the case, but its pretty evident that the technology is going nowhere and anyone who is not incorporating it in some way (workflow or actual end product) is just holding themselves back. The fact that it has opened the doors to allow developers work on ideas that they would not have had to the time to do previously should be celebrated in my opinion. For people who are interested in creative web projects, we live in very exciting times, regardless of the negative noise.

krapp 9mo ago
It isn't just the case that AI generated content tends to be slop, it's that it's intellectually uninteresting and provides little worth discussing on a forum whose entire purpose is intellectually curious conversation. Imagine this was a birdwatching forum and you were arguing that posting LLM generated images of birds allows birdwatchers to work on "watching" birds that they would not have had the time to watch previously, and that this should be celebrated. It's missing the entire point. >The fact that it has opened the doors to allow developers work on ideas that they would not have had to the time to do previously should be celebrated in my opinion. They have the time, they just don't want to put in the effort. Resources and education have never been more widely or freely available. One of the biggest lies people tell about LLMs is that they "democratize creativity." They don't - they commoditize it. You aren't developing the app. You aren't writing the script. You aren't making the art. The billion dollar proprietary black box service you loaded a prompt into shat out an approximation of a product and maybe at best you tweaked some code (more likely just fed it back in to another LLM.) Yes, you wound up with a finished product of dubious quality that you probably don't even understand and can't discuss in any depth. bravo. But it shouldn't surprise anyone that people here - who actually care about the journey and the process rather than simply getting to an MVP as quickly as possible - for the most part aren't going to be impressed and aren't going to care much about it.
- Imustaskforhelp 9mo ago
  I do feel like its much more so the art of cooking than birdwatching (atleast imo) Some people do want their food to be made by a chef sometimes even in front of them explaining them everything and the journey to them makes it worth it For some people, it can be that the food that they buy from a shop nearby in front of them is good For some, they want the cooking to be done out of their visible range and just want the food served And the last categories probably the ones who will go to supermarkets to get some chips with 3 months expiry date created in a large factory Hackernews to me feels like its 1 & 2 for me. What you describe as birdwatching is much rather only 1st category to me. Like, in the 1st category people want to eat the special dish created by the trials of countless nights and passion On the 2nd category, you are still witnessing a person but they are probably following a much more experience and in the start by say following tutorials or watching recipes. But they are much more democratized I don't know, I don't go in hotels that much so I don't know the 1st part and how possible it is but I do go to 2nd category and I go around my local shops and see people make french fries and chat with them and ask how long have they been here and making french fries etc., they are willing to tell you everything. I do feel like "AI slop" is kind of the 2nd category because like I am not only just interested in whether or not this recipe's entirely unique but I am rather interested in everything around it, what infrastructure choices they make and what coding tech stack they used & are they passionate about the project or what did they end up creating the prototype for. the 3rd part is what I despise where think burger king or mcdonald etc. where the recipes come from franchise and everything comes and the person just serves you. All decisions are bland (think yet another nextjs app). I don't think much of us want HN to be a supermarket or another macD but do we want this to be a cooking show where chefs build the whole recipes to serve us or do we want it to be a lot of neighbourly shops who have friendly people working who will chat with us and we can eat the french fries they make :) (I love the french fries my neighbourhood makes, they are really spicy combine them with spring rolls and some drink or a good yogurt and it feels soo good lol, plus all of this costs less a 1$ or 2$ at most 3$ where I live, I can go and eat without worrying at all about the cost of eating outside usually if picked from the right shops) Although I have never had the opportunity of seeing a chef cook his own dish in such sense but I can imagine that its great & I would like them as well but to me hackiness includes both, them hacking around with recipes to build something nice and the other hacking around with infrastructure like what logo and how close they are to me and even things like pricing imo. I consider the latter to be like much more reachable than the former and to me I don't know how to explain but the idea of hackiness isn't just some clean final product, its also like just building something and gluing things and its messy too. And I love both of them both clean and messy hacking, I just want to know how much an human has been in the loop, where'd they get the idea and what they implemented in etc. I don't know I live in a small city connected to a large city and I feel like I enjoy much more the messy aspects whereas people from large cities are much more likely to want to go to a large hotel for a dine or have a more artisanal taste. So I have seen my fair share of people who want both. Right now, I feel like it honestly depends on how large or small community you feel HN is. If you consider HN to be too large aka you find it less likely to trust projects who get to here in the first place, you would want an artisanal person to build things for hand for you to enjoy But if you consider HN to be small and there is just this idea of treating each person more individually and just this unwritten rule of law which generates more trust and faith in trying out new options even if they are hacky as long as I can sort of trust them decently and they are still involved in the loop and not completely 100% detached unlike say a large conglomerate franchise imo Honestly I want HN to be a small place and not a large place but I feel like even HN is a large place for me but the problem with completely small places or building one's own is that you don't get visitors at all. We probably need something in between HN and no social media promotions (only via word of mouth) imo. I have been trying to replace HN for quite a while now for something more nicher but still well connected place as well. Honestly have thought multiple times to create an HN alternative with the same UI and everything and same rules but just smaller and probably federated (ironically Gonna have it be vibe coded in single main.go golang :) hosting on a cheap recurring deal netcup server I got) but moderation of >0 people is hard and getting >0 people interested in a new website is hard as well which is precisely what the main post is about.
Imustaskforhelp 9mo ago
I think I agree with you & I follow a similar trend but the creative part is doing a lot of heavy work. Like, If we are using AI to build yet another next.js App, then perhaps its me but I personally find it doesn't satisfy my curiosity. I want to see projects be written in niche languages and new languages and AI can be perfect for that. See which language you like the most and each language/idea has their own tradeoffs but just follow what you want and now you are less likely to be stuck. Of course though sometimes I create a python script first (I recently made a code where your png's/jpg's can automatically be updated in a slideshow like thing which I built with a custom youtube screenshot and I watched some lectures, pressed shift A and it can now have slides and I compile them to pdf and built another where I can have a pdf and then I can have a laser tool on pdf's because I love to study with notes where I can draw random things on page like whiteboard and erase them without hurting the text below for understanding._ I think converted that python tkinkter gui to fyne golang out of curiosity yesterday and I feel both tools are great fwiw I think my point here is I'd much rather be willing to see some niche projects for ourselves and have it first meaningfully improve our lives & just experiment. I don't know not much for monetary gains but out of curiosity and seeing how far you can and cannot go. Just build things you like/want to exist in any way that you like till AI's so damn subsidized (I don't use any agents, just claude web's free version or aistudio for free, my projects are usually around 600-700locs so I can ask it to regenerate it completely and I am fine with much of it, Perhaps I just follow KISS to its core) You would find it funny but the problem's not in code generation but dependency management etc. atleast to me which is why I love uv scripts a lot but sometimes especially chatgpt doesn't know about uv script feature or hallucinates it hard on their web version so If I am building python applets I would paste uv script's complete docs to the LLM website I would be using.
koakuma-chan 9mo ago
> Obviously this can be the case More often than not that is the case.
plastic041 9mo ago
> HN users that have settled on the belief that if something is done using AI it is "lazy slop" and it needs to be shunned! Honestly, I used to be one of them. I recently saw a great library (atcute) that changed my opinion. I think most AI sceptics haven't had this experience. They saw AI slops and set their opinion: AI-generated codes are bad. I can't really blame them, though, because there are so many of AI slops. > anyone who is not incorporating it in some way (workflow or actual end product) is just holding themselves back. It's true. However, I think people who create AI slops are worse than those who don't use it. They are diligently making this world a worse place.

krapp 9mo ago

It isn't just the case that AI generated content tends to be slop, it's that it's intellectually uninteresting and provides little worth discussing on a forum whose entire purpose is intellectually curious conversation. Imagine this was a birdwatching forum and you were arguing that posting LLM generated images of birds allows birdwatchers to work on "watching" birds that they would not have had the time to watch previously, and that this should be celebrated. It's missing the entire point. >The fact that it has opened the doors to allow developers work on ideas that they would not have had to the time to do previously should be celebrated in my opinion. They have the time, they just don't want to put in the effort. Resources and education have never been more widely or freely available. One of the biggest lies people tell about LLMs is that they "democratize creativity." They don't - they commoditize it. You aren't developing the app. You aren't writing the script. You aren't making the art. The billion dollar proprietary black box service you loaded a prompt into shat out an approximation of a product and maybe at best you tweaked some code (more likely just fed it back in to another LLM.) Yes, you wound up with a finished product of dubious quality that you probably don't even understand and can't discuss in any depth. bravo. But it shouldn't surprise anyone that people here - who actually care about the journey and the process rather than simply getting to an MVP as quickly as possible - for the most part aren't going to be impressed and aren't going to care much about it.

Imustaskforhelp 9mo ago
I do feel like its much more so the art of cooking than birdwatching (atleast imo) Some people do want their food to be made by a chef sometimes even in front of them explaining them everything and the journey to them makes it worth it For some people, it can be that the food that they buy from a shop nearby in front of them is good For some, they want the cooking to be done out of their visible range and just want the food served And the last categories probably the ones who will go to supermarkets to get some chips with 3 months expiry date created in a large factory Hackernews to me feels like its 1 & 2 for me. What you describe as birdwatching is much rather only 1st category to me. Like, in the 1st category people want to eat the special dish created by the trials of countless nights and passion On the 2nd category, you are still witnessing a person but they are probably following a much more experience and in the start by say following tutorials or watching recipes. But they are much more democratized I don't know, I don't go in hotels that much so I don't know the 1st part and how possible it is but I do go to 2nd category and I go around my local shops and see people make french fries and chat with them and ask how long have they been here and making french fries etc., they are willing to tell you everything. I do feel like "AI slop" is kind of the 2nd category because like I am not only just interested in whether or not this recipe's entirely unique but I am rather interested in everything around it, what infrastructure choices they make and what coding tech stack they used & are they passionate about the project or what did they end up creating the prototype for. the 3rd part is what I despise where think burger king or mcdonald etc. where the recipes come from franchise and everything comes and the person just serves you. All decisions are bland (think yet another nextjs app). I don't think much of us want HN to be a supermarket or another macD but do we want this to be a cooking show where chefs build the whole recipes to serve us or do we want it to be a lot of neighbourly shops who have friendly people working who will chat with us and we can eat the french fries they make :) (I love the french fries my neighbourhood makes, they are really spicy combine them with spring rolls and some drink or a good yogurt and it feels soo good lol, plus all of this costs less a 1$ or 2$ at most 3$ where I live, I can go and eat without worrying at all about the cost of eating outside usually if picked from the right shops) Although I have never had the opportunity of seeing a chef cook his own dish in such sense but I can imagine that its great & I would like them as well but to me hackiness includes both, them hacking around with recipes to build something nice and the other hacking around with infrastructure like what logo and how close they are to me and even things like pricing imo. I consider the latter to be like much more reachable than the former and to me I don't know how to explain but the idea of hackiness isn't just some clean final product, its also like just building something and gluing things and its messy too. And I love both of them both clean and messy hacking, I just want to know how much an human has been in the loop, where'd they get the idea and what they implemented in etc. I don't know I live in a small city connected to a large city and I feel like I enjoy much more the messy aspects whereas people from large cities are much more likely to want to go to a large hotel for a dine or have a more artisanal taste. So I have seen my fair share of people who want both. Right now, I feel like it honestly depends on how large or small community you feel HN is. If you consider HN to be too large aka you find it less likely to trust projects who get to here in the first place, you would want an artisanal person to build things for hand for you to enjoy But if you consider HN to be small and there is just this idea of treating each person more individually and just this unwritten rule of law which generates more trust and faith in trying out new options even if they are hacky as long as I can sort of trust them decently and they are still involved in the loop and not completely 100% detached unlike say a large conglomerate franchise imo Honestly I want HN to be a small place and not a large place but I feel like even HN is a large place for me but the problem with completely small places or building one's own is that you don't get visitors at all. We probably need something in between HN and no social media promotions (only via word of mouth) imo. I have been trying to replace HN for quite a while now for something more nicher but still well connected place as well. Honestly have thought multiple times to create an HN alternative with the same UI and everything and same rules but just smaller and probably federated (ironically Gonna have it be vibe coded in single main.go golang :) hosting on a cheap recurring deal netcup server I got) but moderation of >0 people is hard and getting >0 people interested in a new website is hard as well which is precisely what the main post is about.

Imustaskforhelp 9mo ago

I think I agree with you & I follow a similar trend but the creative part is doing a lot of heavy work. Like, If we are using AI to build yet another next.js App, then perhaps its me but I personally find it doesn't satisfy my curiosity. I want to see projects be written in niche languages and new languages and AI can be perfect for that. See which language you like the most and each language/idea has their own tradeoffs but just follow what you want and now you are less likely to be stuck. Of course though sometimes I create a python script first (I recently made a code where your png's/jpg's can automatically be updated in a slideshow like thing which I built with a custom youtube screenshot and I watched some lectures, pressed shift A and it can now have slides and I compile them to pdf and built another where I can have a pdf and then I can have a laser tool on pdf's because I love to study with notes where I can draw random things on page like whiteboard and erase them without hurting the text below for understanding._ I think converted that python tkinkter gui to fyne golang out of curiosity yesterday and I feel both tools are great fwiw I think my point here is I'd much rather be willing to see some niche projects for ourselves and have it first meaningfully improve our lives & just experiment. I don't know not much for monetary gains but out of curiosity and seeing how far you can and cannot go. Just build things you like/want to exist in any way that you like till AI's so damn subsidized (I don't use any agents, just claude web's free version or aistudio for free, my projects are usually around 600-700locs so I can ask it to regenerate it completely and I am fine with much of it, Perhaps I just follow KISS to its core) You would find it funny but the problem's not in code generation but dependency management etc. atleast to me which is why I love uv scripts a lot but sometimes especially chatgpt doesn't know about uv script feature or hallucinates it hard on their web version so If I am building python applets I would paste uv script's complete docs to the LLM website I would be using.

koakuma-chan 9mo ago

> Obviously this can be the case More often than not that is the case.

plastic041 9mo ago

> HN users that have settled on the belief that if something is done using AI it is "lazy slop" and it needs to be shunned! Honestly, I used to be one of them. I recently saw a great library (atcute) that changed my opinion. I think most AI sceptics haven't had this experience. They saw AI slops and set their opinion: AI-generated codes are bad. I can't really blame them, though, because there are so many of AI slops. > anyone who is not incorporating it in some way (workflow or actual end product) is just holding themselves back. It's true. However, I think people who create AI slops are worse than those who don't use it. They are diligently making this world a worse place.